I've witnessed a dramatic and telling moment twice recently. A few months ago, I joined a panel of luminaries and 300-400 data warehouse professionals in an extended Q&A session. One of the questions was, "How many of you have existing data warehouse or data mart systems?" About 60 to 70 percent of the audience raised their hands. The follow up questions was, "How many of those systems met the expectations of the business?" About half a dozen hands went up. Yesterday, while giving a keynote, I asked the same questions and got about the same response, two hands out of 40 or 50 existing systems. A very insightful attendee then asked, "Which expectations do you mean, the ones we thought we set, or the ones they developed on their own?" Therein lies the rub.

As these informal surveys indicate, data warehouse teams are building systems that are technically successful to some degree, but are dramatically falling short of meeting the expectations of the business, and more than that, the teams are not even sure if they know what those expectations are.

In defense of the data warehouse teams, users have an incredible ability to generate runaway expectations all on their own. This usually starts about 20 minutes after the data warehouse team has defined the scope of the system to be delivered. One user will ask another, "Will the data warehouse have sales detail for all stores?" "Yes, I'm pretty sure that's what they said," will be the reply. Soon another user will ask, "Will the data warehouse have transaction history?" The answer will be "Yes, I just was talking with someone who attended the briefing, and they said that we'd have full store transaction detail, and I think they were going to include it back for a few years." The next question will be, "Will the data warehouse have cost information so we can calculate profit by product?" The answer will be, "I was talking with someone who knew about the briefing, and they said it would have full detail for all global transactions since we started business, and it would only make sense if it had all costs down to the transaction/product/channel/customer level." And so it goes...

In reality, the team had briefed a select, targeted user group on the data warehouse pilot for phase one that would contain data from one store for one product line for one three-month period.

In defense of the users, the data warehouse teams tend to be typically fantastic at technology related tasks, but miserably unprepared for the non-technical aspects of data warehousing, such as politics and communication. A typical scenario is that the team will communicate like crazy right up to the moment of winning budget approval for the project, followed by four to eight months of 18- hour work days spent designing, implementing and finally deploying the system. The only contact with the outside world during this time is the occasional sighting of a team member. This event is usually in the context of a pizza delivery person, a Jolt cola truck or around the coffee machine. Eventually the team reappears from their bunker, proudly clutching their limited scope phase one pilot. Of course, in the intervening vacuum, the users have filled in the blanks, as outlined earlier. What results is a mutual disappointment society. The team is disillusioned by the users' disappointment. The users are frustrated by yet another "they didn't build what we needed/expected" experience.

How do you ensure that you are one of the few to proudly raise your hand in response to, "Did your system meet the expectations of the users?"

1. Develop an "elevator test" scope and mission statement. Spend an hour or so on your nearest elevator practicing delivering a succinct scope and mission statement for your data warehouse/data mart system that you can articulate in the time it takes the elevator doors to close and travel one floor. This crisp statement will help you and the team remain on target and greatly ease casual conversation about the project and what it is trying to accomplish.

2. High frequency communications. Create an e-mail distribution list of the key players on the user side. Send out a status e-mail at least every two weeks. In every status report, include your "elevator test" scope and mission statement.

3. Face-to-face relations. Maintain face-to-face relationships with the key players from the user community. This may come as shocking news, but e- mail is no substitute for direct, human communication. While in these face- to- face communications, repeat your elevator test scope and mission statement at least once. Most importantly, listen to the users. If you realize your project is out of sync with their needs, you must immediately address the issue.

4. Leverage communications/ media. One of the most important, yet neglected parts of any DW/DM project is marketing. A key aspect of marketing is media, and you need to leverage every available media outlet: company newsletters, intranet sites, internal TV, etc. Work with your corporate communications department to educate them about data warehousing and its benefits. You'll find that the easier you can make it for them, the more media coverage you'll receive.

Expectation management is just one example of the myriad of non-technical issues that can sabotage your system's potential success. Invest energy into expectation management, and you'll be rewarded by being one of the lonely few who can proudly wave your hand and proclaim "we met and exceeded the expectations of our users!"

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access